ON THE CAE GWYN CAVE. 



127 



distribution of the surface-deposits might help us to form a judg- 

 ment as to the position of any swallow-holes or lines of soakage into 

 the cave. In this I was not mistaken, and the importance of the 

 question was at once recognized by those who differed from me in 

 the interpretation of the section. 



To fix a point where I considered that the re-sorted surface of the 

 marine drift (itself remanie) attained an appreciable thickness, I 

 pointed out an old hedge against which the head or rmi of the hill 

 had been caught. Of the existence of this old fence I might adduce 

 other evidence, if the photographs I exhibited were not sufficient, 

 as it was removed and a portion of the bank levelled by the present 

 owner less than 10 years ago. 



It is within 16 feet of the upper opening into the cave, and runs 

 much nearer to the rock in which the cave occurs. The last exca- 

 vations cut across this bank, and have shown a depth of 6 feet of 

 surface-soil here. This deposit rises to the JS'.E. and thickens out 

 again at the north-east end of the pit, as indicated in fig. 5, which 



Fig. 5. 



The left -band portion represents the section seen along the north-west face of the 

 cutting in September 1887, and shows the festooning of the upper part of the 

 superficial deposits at the margin of the swallow-hole. The portion included 

 within the asterisks is that shown in the photograph, fig. 1. The right- 

 hand portion is the section along the north-east face as exposed by the 

 excavations carried on in October following. 



(Scale 20 ft. to 1 inch.) 



s. sew 



N. 30° E.N.W. 



a. Head, Rainwash, Eun-of-the-Hill. 



b. Moved drift ; marginal portion of that which slipped into swallow-hole, 



and finally closed the upper opening. 



c. Oave-deposits. 



d. Angular limestone ; broken-down wall and roof of cave. 



e. Marine drift. 



/ Shell-bed in drift. 

 g. Limestone. 



shows the section as seen by General Pitt-Rivers and Dr. John 

 Evans in September. This face was afterwards cut back a little, 

 but still, in October, the section given from a photograph (fig. 1) 

 shows the margin of the looped drift cutting off the shell-bed and 

 the surface-soil falling with it in towards the swallow-hole. The 



